Influence of incorporating different electron-rich thiophene-based units on the photovoltaic properties of isoindigo-based conjugated polymers: An experimental and DFT study

Article abstract of DOI:10.1021/ma401691r

A series of novel donor-acceptor conjugated alternating copolymers based on the isoindigo acceptor moiety have been designed, synthesized, and characterized, in order to explore the potential of isoindigo for efficient donor materials with high photovoltages in solar cells. We have systematically investigated and discussed the effect of combining different electron-rich thiophene-based units on the structural, optical, electronic, and photovoltaic properties of the resulting polymers. Morphological studies and quantum-chemical calculations are carried out to gain insights into the different properties. The power conversion efficiencies (PCEs) of the solar cells based on these polymers are increased step by step by over 3-fold through a rational structural modification. Among them, PBDTA-MIM shows a PCE of 5.4%, which is to our knowledge the best result achieved among isoindigo-based polymers for solar cells combined with PC₆₁BM as the acceptor using the conventional device configuration. Our results further emphasize the use of isoindigo as an effective acceptor unit and highlight the importance of carefully choosing appropriate chemical structure to design efficient donor-acceptor polymers for organic solar cells. In addition, the resulting low optical gaps, the promising PCEs with PC₆₁BM as the acceptor, and the good open-circuit voltages (up to 0.8 V) synergistically demonstrate the potential of this class of polymers as donor materials for bottom subcells in organic tandem solar cells.

Full text of DOI:10.1021/ma401691r

Article
pubs.acs.org/Macromolecules
Influence of Incorporating Different Electron-Rich Thiophene-Based
Units on the Photovoltaic Properties of Isoindigo-Based Conjugated
Polymers: An Experimental and DFT Study
Wenliu Zhuang,*,† Margherita Bolognesi,‡ Mirko Seri,§ Patrik Henriksson,† Desta Gedefaw,†
Renee Kroon,† Markus Jarvid,† Angelica Lundin,† Ergang Wang,† Michele Muccini,∥
and Mats R. Andersson*,†
†Department of Chemical and Biological Engineering, Polymer Technology, Chalmers University of Technology, SE-412 96
Goteborg, Sweden
̈
‡Laboratory MIST E-R, Via P Gobetti 101, 40129 Bologna, Italy
§Istituto per la Sintesi Organica e la Fotoreattivita
Bologna, Italy
̀
(ISOF), Consiglio Nazionale delle Ricerche (CNR), Via P. Gobetti 101, 40129
∥Istituto per lo Studio dei Materiali Nanostrutturati (ISMN), Consiglio Nazionale delle Ricerche (CNR), Via P. Gobetti 101, 40129
Bologna, Italy
S
* Supporting Information
ABSTRACT: A series of novel donor−acceptor conjugated
alternating copolymers based on the isoindigo acceptor moiety
have been designed, synthesized, and characterized, in order to
explore the potential of isoindigo for efficient donor materials with
high photovoltages in solar cells. We have systematically investigated
and discussed the effect of combining different electron-rich
thiophene-based units on the structural, optical, electronic, and
photovoltaic properties of the resulting polymers. Morphological
studies and quantum-chemical calculations are carried out to gain
insights into the different properties. The power conversion
efficiencies (PCEs) of the solar cells based on these polymers are increased step by step by over 3-fold through a rational
structural modification. Among them, PBDTA-MIM shows a PCE of 5.4%, which is to our knowledge the best result achieved
among isoindigo-based polymers for solar cells combined with PC61BM as the acceptor using the conventional device
configuration. Our results further emphasize the use of isoindigo as an effective acceptor unit and highlight the importance of
carefully choosing appropriate chemical structure to design efficient donor−acceptor polymers for organic solar cells. In addition,
the resulting low optical gaps, the promising PCEs with PC61BM as the acceptor, and the good open-circuit voltages (up to 0.8
V) synergistically demonstrate the potential of this class of polymers as donor materials for bottom subcells in organic tandem
solar cells.
optimization of device fabrication (i.e., to control active layer
morphology and interfacial engineering).1a,4
INTRODUCTION
■
Polymer solar cells (PSCs) are a promising technology for
producing clean and renewable energy, based on solution-
processed, lightweight, potentially flexible devices.1 The current
challenges for PSCs are to improve photovoltaic efficiency,
cost-effectiveness, and durability. The state-of-the-art device
architectures of PSCs to date are based on the concept of a bulk
heterojunction (BHJ)2 as the active layer, which typically
consists of a bicontinuous interpenetrating blend of an
electron-donating polymer and an electron-accepting fullerene
derivative. The power conversion efficiency (PCE) of PSCs
during the past 4 years has improved rapidly from 5% to
approximately 10%,1a,3 mainly due to the design and synthesis
of new electron-donating conjugated polymers with desirable
chemical and physical properties (i.e., to control frontier
molecular orbital energies and molecular ordering) and
An efficient method to obtain conjugated polymers is based
on the strategy of donor−acceptor polymers,5 by combining π-
electron-rich (donor) and π-electron-deficient (acceptor)
conjugated moieties into their repeating units. This particular
method allows control of the band gap and energy levels of
conjugated polymers by choosing appropriate donor and
acceptor moieties, with the donor typically determining the
energy level of the highest occupied molecular orbital
(HOMO) and the acceptor, the lowest unoccupied molecular
orbital (LUMO).6 To push the efficiency toward the theoretical
Received: August 12, 2013
Revised: September 29, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
limit in single-junction photovoltaic cells, achieving both a high
short-circuit current density (JSC) and a high open-circuit
voltage (VOC) is crucial, along with a high fill factor (FF). Both
JSC and VOC are influenced by the band gap of the photoactive
material and by the alignment of the HOMO and LUMO
energy levels with respect to those of the acceptor material used
in the active blend. Note that the PCE is not monotonically
correlated to either the band gap of the donor material or the
LUMO-level offset of the donor to the acceptor, but both of
them need to be optimized independently, from a materials
chemist’s point of view.6b,7
We systematically investigate the effect of these different
electron-rich units on the chemical, optical, electronic,
morphological, and photovoltaic properties of the resulting
isoindigo-based conjugated polymers. We find that the PCEs of
the resulting PSCs can be improved by over 3-fold through a
rational structural modification of the donor moiety. Among
the four resulting isoindigo-based polymers, PBDTA-MIM in
combination with PC61BM ([6,6]-phenyl-C61-butyric acid
methyl ester) shows a PCE of 5.4%, which is to the best of
our knowledge the highest value for conventional BHJ solar
cells with isoindigo-based polymers blended with PC61BM.
Considering the relatively low optical gaps (1.5−1.7 eV) of
these polymers, the promising PCE of PC61BM-based devices
along with their resulting good VOC (at least 0.8 V)
synergistically makes this class of polymers suitable candidates
as donor components for the bottom active layer in tandem
solar cells.7b
In the exploration of new electron-donating materials,
semiconductors based on isoindigopossessing a ketopyr-
role-containing core of two lactam rings with a strong electron
affinityhave attracted increasing attention due to their
potential in optoelectronic devices.8 Although being already
known for over a century,9 isoindigo was first introduced only
in 2010 by Reynolds and co-workers as an effective building
unit for new conjugated molecular materials and polymers.8a,b
Our recent studies demonstrated the potential of isoindigo as
an efficient acceptor unit in donor−acceptor polymers for BHJ
solar cells.8d,i,j,l Among these, PTI-1, consisting of alternating
thiophene and isoindigo units, with an optical gap of 1.6 eV,
shows a PCE of 3.0% with a VOC of 0.89 V, JSC of 5.4 mA/cm2,
and FF of 63% when combined with PC71BM ([6,6]-phenyl-
C71-butyric acid methyl ester) as the acceptor in BHJ devices.8d
Extending the conjugation length of the donor unit in PTI-1
from thiophene to a terthiophene derivative resulted in a more
efficient donor polymer, P3TI, with a reduced optical gap (1.5
eV). P3TI shows a PCE of 6.3% in the same device
configuration, with a VOC, JSC, and FF of 0.70 V, 13.1 mA/
cm2, and 69%, respectively.8i Notably, even though a high PCE
was achieved for P3TI, the VOC was not prominent, while for
PTI-1, even though a VOC up to 0.9 V was obtained, the PCE
was moderate. The initial success of isoindigo-based polymers
for organic solar cells8c−e,i,j,l,10 motivated us to further explore
the possibility of improving the photovoltaic performance in
these systems, especially by combining a high PCE and a high
VOC into one single isoindigo-based polymer, as well as to
investigate their structure−property relationships.
RESULTS AND DISCUSSION
■
Material Design, Synthesis, and Structural Character-
ization. Four donor−acceptor alternating copolymers (Chart
1) were synthesized according to Scheme 1. Previous work on
Scheme 1. Synthesis of Monomers and Polymers
In this work we report on the synthesis, characterization, and
structure−property relationships of a series of novel electron
donor polymers (PMIM, PT-MIM, PBDTO-MIM, and
PBDTA-MIM, Chart 1), which combine the same isoindigo
acceptor moiety with different thiophene-based donor units.
Chart 1. Chemical Structures of PMIM, PT-MIM, PBDTO-
MIM, and PBDTA-MIM
isoindigo polymers has shown that extending the conjugation in
the donor moieties from a single thiophene (PTI-1)8d to
terthiophene (P3TI)8i resulted in a decreased VOC by ∼0.2 V,
even though the PCE was more than doubled. In this study, the
bithiophene unit was employed to have an intermediate π-
conjugation length in the donor moiety, resulting in donor−
acceptor polymer PMIM. The introduction of methyl groups
on the bithiophene unit would be expected to give a slightly
twisted backbone, which should increase the band gap and thus
11
the VOC
.
Meanwhile, a beneficial intermolecular stacking
B
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
might still occur by adopting conformations where the two
methyl substituents on the adjacent thiophene units are away
from each other,12 favoring a linear backbone in the solid
state.13 For comparison, a thiophene ring was inserted in
between the two alkylthiophene units to have the same polymer
backbone as P3TI (the best-performing isoindigo-based
polymer reported so far),8i resulting in PT-MIM. To further
extend the π-conjugation length of the donor unit, benzo[1,2-
b:4,5-b′]dithiophene (BDT) was also incorporated. Recently,
BDT has been widely used as an effective electron-rich building
block thanks to its desirable properties such as structural
rigidity, planarity, and favorable interchain π−π stacking, along
with the presence of additional substitution sites for the
incorporation of side chains.4a−h,k,p Here the BDT unit with
two 2-ethylhexyloxy side chains was used to replace the middle
thiophene in the terthiophene unit in PT-MIM, resulting in
PBDTO-MIM. Notably, similar polymers with identical
backbones (PBDT-TIT and PBDT-OIO) showed promising
device performance with a VOC up to 0.79 V, JSC of 7.87 mA/
cm2, FF of 68%, and a PCE of 4.22%.8l Furthermore, it is
known that removal of oxygen from the side chains of the
donor unit can reduce the electron density, to allow for a
deeper HOMO energy level, which in principle should enhance
the VOC of the resulting PSCs.14 Hence, the alkoxy side chains
on the BDT unit were replaced with branched alkyl chains,
resulting in PBDTA-MIM.
extraction with methanol, acetone, ether, hexane, dichloro-
methane, and chloroform successively. The chloroform fraction
was precipitated into methanol, and the final products were
collected by filtration and dried under vacuum. The molar
masses of the four polymers were determined by size-exclusion
chromatography (SEC) relative to uniform polystyrene. The
corresponding number-average degrees of polymerization (Xn)
were also calculated for comparison. In order to obtain high
relative molecular masses, the equimolar feed ratios of the
respective monomers and the polymerization time were
carefully controlled until precipitation of the resulting polymers
began to occur.
PMIM had a number-average molar mass (Mn) of 20.1 kDa
with a molar-mass dispersity (ĐM) of 2.1 (Xn = 22.3). It should
be mentioned that a hot chlorobenzene fraction after
chloroform extraction gave a comparable yield of PMIM, with
a much higher Mn of 62.8 kDa (ĐM = 1.9, Xn = 69.7), which is
not considered for comparison in this work because it was not
processable from common organic solvents at the concen-
trations needed for devices preparation (see also further
discussions about absorption in solution). Note that the
polymerization of PMIM was very fast (15 min at 95 °C).
Therefore, for the other three polymers lower concentrations
were employed to obtain good amounts of soluble polymers
from hot chloroform fraction. Going from PMIM to PT-MIM
resulted in a slightly lower Mn of 13.4 kDa for PT-MIM (ĐM =
1.8, Xn = 13.6), indicating a decreased solubility. Note that both
PMIM and PT-MIM had a lower Xn (and Mn) than their
respective analogues IIDDT and IID-T3 (without any
substituent on the thiophene rings but with slightly longer
side chains on the isoindigo) from the chloroform-soluble
fractions (IIDDT: Mn = 33.7 kDa, ĐM = 5.4, Xn = 34.2; IID-T3:
Mn = 33.8 kDa, ĐM = 3.4, Xn = 36.3).19 Given their same
backbone structures, the solubility can vary because of the
difference in the side chains on the isoindigo units (2-
hexyldecyl in this work and 2-octyldodecyl in the literature) as
well as on the flanking thiophenes (with or without methyl
groups, respectively). In fact, the introduction of methyl groups
on the flanking thiophenes did not guarantee increased molar
masses for the resulting polymers with sufficient solubility, for
example, going from IIDDT to IIDDT-Me (Mn = 20.1 kDa, ĐM
= 3.9, Xn = 17.2) in the same literature,19 probably due to an
increased rotational barrier between adjacent thiophene rings
(with head-to-head linkages in the presence of substituents).12
Also, different polymerization conditions can further explain the
observed variation of the relative molecular masses. By
replacing the unsubstituted thiophene with a BDT unit with
alkoxy or alkyl branched side chains, a dramatic improvement
in terms of solubility and molar masses of the polymers
occurred (PBDTO-MIM: Mn = 63.5 kDa, ĐM = 3.0, Xn = 47.2;
PBDTA-MIM: Mn = 69.3 kDa, ĐM = 4.0, Xn = 47.7).20 Finally,
all the polymers in this study were washed with the same
solvents by Soxhlet extraction as mentioned before, obtained
from the chloroform fraction, and were soluble in common
solvents such as chloroform, chlorobenzene, o-dichlorobenzene
(DCB), etc. The comparison of these polymers in this work is
therefore more based on a compromise between sufficient
solubility and high molar masses, both of which are relevant
properties to the solution-processing conditions and film
morphology.21
The synthesis of the monomers is outlined in Scheme 1.
Monomer 5,5′-bis(trimethylstannyl)-3,3′-dimethyl-2,2′-bithio-
phene (M1) was readily prepared through a three-step
procedure starting from commercially available 3-methylthio-
phene (1). First, bromination of 1 with N-bromosuccinimide
(NBS) in acetic acid gave 2-bromo-3-methylthiophene (2) in
82% yield, which was followed by a nickel-catalyzed Kumada
coupling to afford 3,3′-dimethyl-2,2′-bithiophene (3) in 79%
yield. Finally, lithiation with n-butyllithium and stannylation
with trimethyltin chloride resulted in monomer M1 in
reasonable yield. Monomer 2,6-bis(trimethylstannyl)-4,8-bis-
(3-butylnonyl)benzo[1,2-b:4,5-b′]dithiophene (M4)4e,f was
prepared according to modified literature procedures.14b,15
Monomer 6,6′-bis(5-bromo-4-methylthiophen-2-yl)-N,N′-bis-
(2-hexyldecyl)isoindigo (M6) was prepared from N,N′-bis(2-
hexyldecyl)-6,6′-dibromoisoindigo (M5) following the proce-
dure described in our previous work,8d,i except tributyl(4-
methylthiophen-2-yl)stannane (9)16 was used instead of
tributyl(4-octylthiophen-2-yl)stannane. Compound 9 was
prepared from 3-methylthiophene (1) by lithiation with n-
butyllithium at −78 °C and stannylation with tributyltin
chloride in quantitative conversion but with a regioselectivity17
of 4:1 over the isomer tributyl(3-methylthiophen-2-yl)stannane
1
as determined by H NMR. The use of this product without
further purification in the Stille coupling with dibromide M5
turned out to give a satisfactory selectivity, which is probably
due to the use of compound 9 in large excess as well as its
inferior steric hindrance effect in comparison with the
unwanted regioisomer in the reaction. N,N′-Bis(2-hexyldec-
yl)-6,6′-bis(4-methylthiophen-2-yl)isoindigo (10) was obtained
in above 90% yield based on compound M5, and subsequent
bromination with NBS in THF afforded monomer M6 in 75%
yield after column chromatography and recrystallization.
All the polymers were synthesized via Stille coupling
reaction18 from the corresponding monomers as shown in
Scheme 1. After the reaction mixture was precipitated into
methanol, the polymers were collected and washed via Soxhlet
Optical and Electrochemical Properties. The normal-
ized UV−vis absorption spectra both in DCB solution and in
pristine films of the four polymers are shown in Figure 1. We
C
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Chart 2. Chemical Structures of Some Isoindigo-Based
Polymers in the Literature
Figure 2. Energy levels estimated from DFT at the B3LYP/6-31G(d)
level and from SWV and optical gaps from film absorption.
well as the reduction in polymer chain length suffering from the
decreased solubilizing side chain density. In contrast, the
analogous P3TI, with long alkyl side chains on the flanking
thiophenes (i.e., octyl vs methyl),8i shows a smaller optical gap
(1.50 eV for P3TI vs 1.57 eV for PT-MIM), presumably
because P3TI has a polymer chain ca. 3 times as long as PT-
MIM, as implied by the obtained molar masses (Mn over 43
kDa for P3TI and only 13 kDa for PT-MIM) and degrees of
polymerization (Xn of 36.4 for P3TI vs 13.6 for PT-MIM). This
corresponds to a difference of around 38 nm in their optical
gaps and indicates PT-MIM may have a conjugation path
length shorter than its effective conjugation length. Further
comparison with another analogue IID-T3 (Eg = 1.58 eV, Xn =
36.3)19 implies that PT-MIM (Eg = 1.57 eV, Xn = 13.6) has a
better conjugation pathway to narrow the band gap than IID-
T3, which may benefit from the introduction of methyl groups
on the flanking thiophenes, favoring a good interchain packing
or a linear backbone in the solid state.13 These comparisons can
explain the observed blue-shift going from PMIM to PT-MIM.
Insertion of BDT units instead of thiophene in the polymer
backbone gave an even larger blue-shift in the solid state (up to
around 60 nm) as compared with PMIM, regardless of alkoxy
(PBDTO-MIM) or alkyl (PBDTA-MIM) side chains attached.
PBDTO-MIM and PBDTA-MIM have a very similar Xn of
around 47 and optical gap (only 3 nm blue-shifted from
PBDTO-MIM to PBDTA-MIM), indicating that the two
polymers have similar effective conjugation lengths.
first considered the optical properties from film absorption
(Figure 1A). All of these polymers show dominant absorption
bands in the long-wavelength side (around 600−660 nm),
which can be assigned to the intramolecular HOMO → LUMO
transitions from the ground state (S0) to the excited state
(S1).22 The appearance of extra peaks or vibrational shoulders
in the long-wavelength side (ca. 670−750 nm) of the
absorption spectra for all of these polymers might originate
from intermolecular transitions.23 According to the absorption
spectra of the films (Figure 1A), the optical gaps of the
polymers, deduced from their absorption edges, are 1.53 eV for
PMIM, 1.57 eV for PT-MIM, and 1.66 eV for PBDTO-MIM
and PBDTA-MIM (Figure 2).
In the solid state, the optical gap of PMIM is slightly lower
than that of the analogous IIDDT (Eg = 1.59 eV), even though
PMIM has a shorter polymer chain than IIDDT as can be seen
from their Xn’s (22.3 vs 34.2).19 This indicates that PMIM has a
better conjugation pathway, probably due to a favored
conformation with the two methyl groups on the bithiophene
unit away from each other, along with an increased inter-ring
rotational barrier, favoring a linear polymer backbone and an
improved interchain packing in the solid state.12,13 PMIM can
absorb light up to ca. 810 nm in the solid state, which would be
beneficial for achieving high photocurrents. Going from PMIM
to PT-MIM resulted in an unexpected blue-shift of ca. 20 nm in
the solid state, while the insertion of thiophene unit should
have decreased the band gap for PT-MIM as compared with
PMIM. This is probably due to increased backbone flexibility in
the absence of head-to-head linkages in the donor moiety as
We also compared the UV−vis absorption of the four
polymers in DCB solution (Figure 1B). The solutions were
prepared in a concentration dilute enough so that the as-
Figure 1. UV−vis absorption spectra of the four isoindigo-based polymers (A) in the solid state and (B) in DCB solution.
D
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
recorded absorbance maxima at the low-energy side ap-
proached unity (corresponding to nearly 90% absorption of
the incident light at most). For PMIM and PT-MIM, the
absorption spectra show a red-shift going from solution to
films, indicating that there are aggregations and/or orderly π−π
stacking formed in the solid state. For PMIM we did not
observe an obvious blue-shifted absorption from film to
solution, which was reported for the aforementioned IIDDT-
Me.19 Instead we observed, even in the dilute solution, a lower
energy shoulder, the relative intensity of which could be
decreased with heating, but did not vanish with further heating
(Figure S1, Supporting Information), indicating a strong
aggregation tendency for PMIM. The difference between
PMIM and IIDDT-Me may be due to the fact that longer
branched side chains were employed, but shorter polymer main
chains were obtained for IIDDT-Me, resulting in a higher
solubility, than for PMIM (see above). The strong aggregation
tendency of PMIM can also be accountable for the exclusion in
this study of the aforementioned much higher molar mass
fraction suffering from poor solubility and poor film-forming
ability. However, for the two BDT-based polymers, only a slight
difference has been observed between the absorption spectra in
solution and in thin films, probably indicating a more rigid
planar conformation, which is a typical phenomenon in the
reported isoindigo-based polymers.8b,c,l,19,24 It has been shown
that there is no obvious aggregation in solution or in thin films
for polymers based on PBDTO-MIM analogues with either
much longer side chains (PBDT-OIO) or no substituents
(PBDT-TIT) on the thiophene spacer due to the bulky
branched side chains on the acceptor units.8l Considering the
chemical structure and the above-obtained molar masses for
PBDTO-MIM and PBDTA-MIM, these two BDT-based
polymers should show absorption coefficients comparable to
PBDT-OIO and PBDT-TIT.8l
the variation in the LUMO energy level therefore mainly
depends on the nature of the donor units. The estimated
LUMO energy levels of the four polymers follow the order
PBDTO-MIM (−3.91 eV) < PMIM (−3.84 eV) < PT-MIM
(−3.83 eV) < PBDTA-MIM (−3.81 eV). For comparison,
energy levels deduced from the first reduction and oxidation
peak potentials of PC61BM are also included in Figure 2. The
LUMO-level offsets between the polymers and PCBM (LUMO
levels of −4.03 and −4.13 eV for PC61BM and PC71BM,
respectively),25 which are considered to provide the driving
force for exciton dissociation as well as to prevent
recombination of photogenerated charges, are small (<0.2
eV) but may still be sufficient.10c
Theoretical Calculations. Quantum-chemical calculations
by the density functional theory (DFT) at the B3LYP/6-
31G(d) level26 using the Gaussian 09 program27 were
employed to demonstrate the electronic structure of these
polymers. For computational simplification, the alkyl side
chains on the isoindigo unit were replaced with methyl groups,
the alkoxy side chains on the BDT unit were replaced with
methoxy groups, and the branched alkyl side chains on the
BDT were replaced with the equivalent ethyl side chains; and
the backbones were simplified to two repeating units. For
PMIM, which has the shortest conjugation path length, we also
extended the computation to three repeating units. All of the
optimized molecular structures are in a stable local minimum of
the ground state potential energy surface, and analytical second
derivatives of the Hessian matrix were calculated at the same
level of theory to ensure this. The optimized molecular
geometries of the models and their calculated frontier orbitals
are depicted in Table S1 (Supporting Information). The
dihedral angles are given in Table 1. Without insertion of
additional donor units, PMIM gives the most twisted (ca. 50°)
optimized molecular geometry as can be seen from Table S1.
The other three polymers can have a more coplanar
conformation than PMIM. The torsion angles between the
flanking methylthiophene and inserted donor units are more or
The HOMO and LUMO energy levels of the polymers
(Figure 2, blue lines) were deduced from their respective
ionization potentials and electron affinities estimated by square-
wave voltammetry (SWV) (Figure S2, Supporting Information)
from the first oxidation and reduction peak potentials,
respectively. All the polymers exhibit deeply low-lying
HOMO levels, in the order: PBDTA-MIM (−5.94 eV) <
PMIM (−5.93 eV) < PBDTO-MIM (−5.87 eV) < PT-MIM
(−5.78 eV).6b Going from PMIM to PT-MIM slightly raised
the HOMO level, and thereby gave a lower offset to the LUMO
Table 1. Dihedral Angles (deg) along the Conjugated
Backbone for the Optimized Molecular Geometries
Obtained by DFT Evaluated at the B3LYP/6-31G(d) Level
energy level of PCBM, which is regarded to be the maximum
6b,8l
value for VOC
.
It is noteworthy that this was accompanied
by a slightly larger optical gap for PT-MIM as discussed above.
On the other hand, the incorporation of BDT units resulted in
relatively deeper HOMO energy levels (−5.87 eV for PBDTO-
MIM and −5.94 eV for PBDTA-MIM) than that of PT-MIM
(−5.78 eV) (and P3TI), comparable to that of PMIM (−5.93
eV). Note that by replacing the alkoxy side chains with the less
electron-donating alkyl chains (going from PBDTO-MIM to
PBDTA-MIM), the HOMO energy level was lowered by 0.07
eV, which would be expected to result in a higher VOC for PSCs
based on PBDTA-MIM than that for PBDTO-MIM.
The LUMO levels of the polymers are very low-lying, which
can be ascribed to their same isoindigo units. This is because
the LUMO levels of donor−acceptor polymers are mainly
governed by the electron-deficient units,6b,e and unlike the
HOMO levels, the LUMO levels are less affected by the
conjugation length of the polymers. Given that the contribution
of the isoindigo moiety to the LUMO level should be similar,
aAll the values are given as positive angles from planarity with all the
adjacent thiophenes adopting anti-conformations. Evaluated with the
terthiophene in a helical conformation, the value in the parentheses
corresponding to a zigzag conformation.
b
E
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Table 2. Photovoltaic Parameters, under Standard Illumination, of the BHJ PSCs and Corresponding Charge Mobilities
a
c
polymer:acceptor ratio
(w/w)
thickness
(nm)
DIO
annealing T
(°C)
JSC
VOC
FF
μh
b
solvent
CHCl3
(% v/v)
(mA cm−2
)
(V)
(%)
PCE (%)
(cm2 V−1 s−1
)
PMIM:PC61BM (1:1)
85
110
130
2
1
3
90
6.24
10.41
10.09
0.73
0.71
0.73
36
43
59
1.64 (1.5)
3.14 (3.1)
4.36 (3.9)
7.4 × 10−5
1.0 × 10−4
1.5 × 10−4
PT-MIM:PC61BM (1:1)
CHCl3
PBDTO-MIM:PC61BM
(1:1.5)
CHCl3:DCB
(1:1)
110
PBDTA-MIM:PC61BM
(1:1.5)
CHCl3:DCB
(1:1)
140
3
10.20
0.80
65
5.36 (5.1)
2.7 × 10−4
PMIM:PC71BM (1:1)
CHCl3
CHCl3
85
95
2
1
3
90
6.54
10.56
10.70
0.71
0.70
0.73
33
48
64
1.54 (1.5)
3.55 (3.4)
4.98 (4.9)
6.0 × 10−5
1.1 × 10−4
2.9 × 10−4
PT-MIM:PC71BM (1:1)
PBDTO-MIM:PC71BM
(1:1.5)
CHCl3:DCB
(1:1)
125
110
PBDTA-MIM:PC71BM
(1:1.5)
CHCl3:DCB
(1:1)
135
3
9.47
0.80
61
4.60 (4.4)
1.2 × 10−4
a
b
Annealing time: 10 min. Power conversion efficiencies of the best devices, with average values based on over six devices given in parentheses.
c
Hole mobility estimated by space charge limited current (SCLC) method.
less the same, around 20°. For PT-MIM, it is also possible to
give a flat angle of θ3 similar to the BDT-based polymers, given
the rigidity and symmetry of the inserted units. As the side
chains on the flanking thiophene are very short (just methyl
groups for the synthesized PT-MIM), they do not force the
terthiophene units to adopt conformations that place the side
chains away from each other.11 The high planarity of the
backbones in these three conjugation-extended polymeric
systems, which was also shown for PBDT-TIT in our recent
modeling work,28 could favor the π−π stacking interactions that
occur among the polymer backbones in the solid state, which
could in turn enhance the charge mobility (see mobility
measurements in Table 2).11
DFT showed very similar LUMOs for PMIM and PT-MIM,
which are rather spatially delocalized, although the wave
functions reside mainly on isoindigo units. Instead, for the
BDT-based polymers, the LUMOs are more localized on the
isoindigo units. The calculated LUMO levels (included in
Figure 2) follow the order PBDTO-MIM (−2.87 eV) <
PBDTA-MIM (−2.85 eV) < PT-MIM (−2.83 eV) < PMIM
(−2.81 eV). Both DFT and SWV showed that PBDTO-MIM
has the lowest LUMO level. The other three polymers showed
a slightly different order of LUMO levels evaluated by DFT
calculations compared to the electrochemical measurements;
however, the differences are very small (0.04 eV for DFT vs
0.03 eV for SWV). This also emphasizes that the LUMO level
of the molecules is mainly governed by the acceptor units, with
a small contribution from the donor units.29 We also looked at
the wave functions of the isolated donor units and found that
the LUMOs of the resulting donor−acceptor molecules are the
bonding linear combination from the LUMO of the isoindigo
segment MIM with the respective LUMOs of the inserted
donor units (Table S2). And the LUMO levels of the inserted
donor units follow the order thiophene > BDTA > BDTO. As a
consequence, the resulting LUMO levels followed the same
order even though the difference is small as can be seen Figure
2, where PMIM has the highest LUMO level over the other
three polymers evaluated by DFT based on the same number of
repeating units. Similar phenomena were observed for the
HOMOs of the polymers, but from the antibonding linear
combination from the HOMOs of the corresponding separated
segments (Table S2). Therefore, PMIM has the lowest HOMO
level as confirmed by DFT, while the difference between the
HOMOs of the other three polymers is even smaller. For all the
four polymers, the HOMOs are delocalized along the entire
conjugated backbones (Table S1).
We also have to keep in mind the influence of different
conjugation path lengths. By limiting the models to the same
numbers of repeating units, systematic error cannot be avoided
from such a low calculation level of oligomers. The conjugation
path lengths in the calculations are quite often shorter than the
effective conjugation lengths, and the change from incorpo-
ration of different donor units will definitely result in the
variation in the conjugation path lengths. This will influence the
extrapolation from the comparison of oligomers to that of
polymers. For this reason, we also extended the DFT
calculation for PMIM to three repeating units to have a similar
conjugation path length to the BDT-based oligomers. The
DFT-computed HOMO and LUMO energy levels are −5.06
and −2.86 eV, respectively. By increasing the number of
repeating units from two to three, we see a ca. 0.05 eV decrease
in the LUMO level and 0.02 eV in the HOMO level. Indeed,
this will result in slightly different orders of the values, but we
can further confirm that the LUMO levels are governed by the
acceptor units and that PMIM has the lowest HOMO level and
the largest LUMO−HOMO energy gap among the four
isoindigo-based polymers in the gas phase.
We also have to mention that the removal of oxygen from
the side chains normally results in a deeper HOMO level, but
this was not observed by DFT for the two BDT-based polymers
with the same conjugated backbone. This turns out to be
accountable for the mismatch of the trends in HOMO levels
from DFT in comparison with electrochemical measurements.
On the other hand, given the fact that the DFT calculations are
based on the gas phase while the experimental results are from
solid states or solution, variations in the trend of the energy
levels between experimental measurements and theoretical
calculations may occur, but this may give a hint to extrapolate
the results from DFT to understand the difference between
different physical states.
Furthermore, we also performed single-point calculations at
the B3LYP/6-31G(d) level of theory on the planar geometries
which were derived from the corresponding optimized
geometries at the same level of theory by forcing the dihedral
angles along the conjugated backbones to be planar. The
energy difference of the forced planar geometry and the
optimized geometry is summarized in Table S3. This was done
to gain more insights about how much energy may be required
to planarize the polymer backbone in order to optimize solid
F
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 3. J−V curves of the optimized BHJ PSCs from the four isoindigo-based polymers with (A) PC61BM and (B) PC71BM as the acceptor.
state packing.11 It has to be noted that solid state packing does
not necessarily cause the polymer backbones to be totally
planar, and therefore the backbone planarization energy
calculated here can be regarded as the upper limit of the
energy required to planarize the backbone upon solid state
packing. Because the substitutions on the thiophenes are only
methyl groups, it is possible to have a more planar backbone
upon solid state packing for PMIM. It is found that due to the
steric hindrance in the presence of head-to-head linkages
between the adjacent substituted thiophene rings in the donor
moiety, the backbone planarization energy for PMIM (10.2
kcal/mol for the two repeating unit and 19.3 kcal/mol for the
three repeating unit) is much larger than the other three
polymers (4.5, 4.0, and 4.8 kcal/mol for PT-MIM, PBDTO-
MIM, and PBDTA-MIM, respectively, based on two repeating
units), as can be seen from Table S3. Considering the
temperature and energy gain upon stacking, it is reasonable
that the planarization energies can be overcome. This will result
in a possibility that in the solid state the LUMO−HOMO
energy gaps of the four isoindigo-based polymers may follow
the same trend as the optical gaps (increasing in the order of
PMIM < PT-MIM < PBDTO-MIM ≈ PBDTA-MIM).
Figure S3). We find that PMIM indeed has a comparable
oscillator strength to the other three polymers and still the
maximum lowest excitation energy, consistent with the largest
DFT-derived LUMO−HOMO energy gap among the four
polymers. This polymer, however, has the smallest optical gap
among the four isoindigo-based polymers deduced from their
film absorption edges. We therefore postulate that the
extension of conjugation for PMIM going from the gas phase
to the solid state may be due to a more coplanar conformation
favored by intermolecular interactions.
We have further looked at the SWV measurements to
understand the mismatch of the trend of the measured and
calculated HOMO and LUMO levels. Energy levels deduced
from the onsets of the first oxidation and reduction potentials
were therefore included in Figure 2 for comparison. The trend
of LUMO levels from these onsets is consistent with the
calculated trend, which is PBDTO-MIM < PBDTA-MIM < PT-
MIM ≈ PMIM. The measured HOMO levels and band gaps
follow the trends from DFT calculations except for PBDTA-
MIM due to the aforementioned reasons.
Photovoltaic Performance. In order to investigate the
effect of incorporating different electron-rich thiophene-based
units into the polymer backbones on the photovoltaic
properties, BHJ PSCs based on the new donor polymers (D)
combined with PC61BM (or PC71BM) as the acceptor (A) were
fabricated using a conventional device configuration (glass/
ITO/PEDOT:PSS/active layer/LiF/Al). Further details for the
device fabrication and characterization are given in the
experimental section. Table 2 and Table S5 summarize the
photovoltaic responses of optimized devices. Representative
current density−voltage (J−V) plots, measured under standard
illumination (AM1.5G, 100 mW cm−2), are shown in Figure 3.
To better evaluate the structure−property relationships of
the different isoindigo-based conjugated polymers, first PSC
optimization was carried out by using PC61BM as the acceptor
material (Table 2, Table S5, and Figure 3A). The optimal D:A
ratio (w/w) was found to be 1:1 for PMIM and PT-MIM or
1:1.5 for PBDTO-MIM and PBDTA-MIM. The slight
difference could be ascribed to the different relative molecular
masses, solubility, and interchain packing tendencies of the two
series of polymers.32 Upon increasing or decreasing the D:A
ratios of the active blends from their optimized values, a
significant drop in JSC, VOC, or FF was observed, resulting in
lower PCEs (Table S5).
Time-dependent DFT (TD-DFT) calculations30 were
performed to assess the excited-state vertical transition energies
and oscillator strengths based on the optimized molecular
geometries at the same level of theory. We find that for all the
polymers all the transitions in the visible region correspond to
π−π* transitions. In particular, the lowest excitations are all
dominated by the HOMO → LUMO transitions as can be seen
from Figure S3 and Table S4.31 The higher energy excitation
bands are also of π−π* character by checking the orbitals in the
major contribution of transitions given in Table S4 and are
primarily from local excitations. The BDT-based polymers
show comparable lowest excitation energies, oscillator
strengths, and high absorption coefficients, with the same
conjugation path length. PT-MIM also showed comparable
absorption properties to BDT-based polymers considering that
PT-MIM in our calculation has a slightly shorter conjugation
path length (22 double bonds for PT-MIM vs 26 for PBDTO-
MIM and PBDTA-MIM). PMIM has the shortest conjugation
path length in the two-repeating-unit models (18 double
bonds), which may render a similar absorption pattern to the
other polymers when computed with comparable conjugation
path lengths. For this reason, we also performed TD-DFT for
PMIM with three repeating units based on the optimized
molecular geometry at the same level of theory (Table S4 and
Several efforts to optimize the thin-film nanomorphology of
the polymer:PC61BM devices were carried out (Table S5) using
G
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
different solvents (e.g., chloroform, chlorobenzene, or o-
dicholobenzene) and thermal annealing (in the range 80−140
°C and for different times). Alternative approaches to enhance
solar cell performance included the use of (i) 1,8-diiodooctane
(DIO) as a processing solvent additive and (ii) PC71BM as the
acceptor.
indicates that the introduction of methyl groups on the
flanking thiophenes is indeed a simple and effective alternative
synthetic strategy without sacrificing the photovoltaic perform-
ance or suffering the synthetic difficulty for example to make
PBDT-TIT.8l This strategy also allows one to design and
synthesize new isoindigo-based polymers that combine many
other different donor moieties in order to further explore
possibilities of improving the photovoltaic performance without
having to stay with the low-yielding synthesis of the dibromide
monomer (i.e., N,N′-disubstituted-6,6′-bis(5-bromothiophen-2-
yl)isoindigo). It is known that having a simple, high-yielding
synthetic route that will reduce materials cost is of critical
importance for the future mass production of PSCs.
The open-circuit voltages of optimized BHJ PSCs for the
four polymers combined with PC61BM (Table 2) track the
progression VOC(PT-MIM) < VOC(PMIM) ≈ VOC(PBDTO-
MIM) < VOC(PBDTA-MIM) rather than VOC(PT-MIM) <
VOC(PBDTO-MIM) < VOC(PMIM) < VOC(PBDTA-MIM) as
expected from E(LUMOA) − E(HOMOD) theoretical
considerations.6b,34 The larger loss in VOC for PMIM, despite
its low-lying HOMO level, could be ascribed to its poor D:A
phase segregation (see AFM images, Figure 5) and low exciton
From the device based on PMIM:PC61BM (1:1) fabricated
from chloroform solution, a PCE of 0.09% with VOC = 0.78 V,
JSC = 0.55 mA/cm2, and FF = 22% was first recorded (Table
S5). Since DIO was shown to be an effective processing
additive for improving the morphology of the active layer4q,33
and thereby enhancing the photovoltaic performance of
isoindigo-based polymers in our previous studies,8d,i 2% (v/v)
of DIO was added to the solution of the active material in order
to further optimize the PV performance. As a result, the JSC of
the device was improved up to 3.48 mA/cm2, resulting in a
PCE of 0.59%. In combination with thermal annealing at 90 °C
for 10 min the PCE of the PMIM:PC61BM-based device was
further improved up to 1.64%, with VOC = 0.73 V, JSC = 6.24
mA/cm2, and FF = 36% (Table 2). Compared to PMIM, PT-
MIM:PC61BM-based solar cells, processed with DIO, exhibit an
almost doubled PCE (3.14%) mainly due to an increased
photocurrent (JSC = 10.41 mA/cm2). This correlates well with
the higher hole mobility (μh) measured for PT-MIM than that
for PMIM (mobilities estimated by space charge limited current
(SCLC), Table 2). However, the VOC and the FF (0.71 V and
43%, respectively) still indicate some morphological limitations
in the optimization of the performances of the PT-
MIM:PC61BM-based device. Additional thermal treatment
(110 °C, 10 min, Table S5) mainly led to a drop in JSC, thus
resulting in a lower PCE.
For PBDTO-MIM and PBDTA-MIM a mixture of chloro-
form and ODCB (1:1 v/v) was used as solvent, since higher
relative molecular masses could be obtained for these polymers,
implying a higher solubility in chlorinated aromatic solvents.
The best-performing BHJ solar cells were obtained again when
processed with DIO as the solvent additive. Because of their
similar molecular structure, with the exception for the branched
side chains (alkoxy or alkyl) on the BDT unit, both PBDTO-
MIM:PC61BM- and PBDTA-MIM:PC61BM-based PSCs exhibit
high PCEs (4.36% and 5.36%, respectively). In particular, the
devices afford similarly high JSC (∼10 mA/cm2), in agreement
with the high charge mobilities of the two polymers (Table 2).
On the other hand, the optimized PBDTO-MIM:PC61BM solar
cell showed a relatively lower VOC (0.73 V vs 0.80 V) than the
PBDTA-MIM-based one, in agreement with the different
HOMO energy levels measured for the two polymers (Figure
2). Despite the slight difference in the FF of the two devices, it
is worth noting that the high JSC (and PCE) obtained for the
two devices mainly suggest good charge transport properties of
both photoactive materials, as a result of the high polymer
charge mobility and an optimal D:A self-organization in the
solid state (see also Morphological Studies section).
dissociation and charge transportall factors that are known to
35
influence the VOC
.
PT-MIM-, PBDTO-MIM-, and PBDTA-
MIM-based solar cells, in agreement with the electrochemically
derived E(LUMOA) − E(HOMOD) offset and in combination
with good film-forming properties, show smaller losses in VOC
,
achieving relatively high VOC (up to 0.8 V).
Therefore, from this set of optimized BHJ devices (based on
PC61BM), some effects of the chemical modifications of the
polymer structures on the corresponding photovoltaic proper-
ties have been evidenced. The insertion of the thiophene
moiety into the backbone, going from PMIM to PT-MIM,
resulted in a decreased VOC due to the raised HOMO level, but
an almost doubled PCE was achieved coupled with a greatly
increased JSC due to the increased hole mobility. In addition,
the replacement of a thiophene ring with a BDT moiety
increased the VOC and FF without losses in JSC, leading to an
enhanced PCE. More importantly, when alkoxy side chains on
the BDT unit were replaced with analogous alkyl chains, the
VOC was further improved due to the deepening in the HOMO
level, and the FF was further improved as well probably thanks
to a modified morphological D:A blend arrangement. As a
result, PBDTA-MIM, blended with PC61BM, stands out as the
best-performing photoactive material among these polymers,
with a VOC of 0.80 V, JSC of 10.20 mA/cm2, FF of 65%, and a
PCE of 5.36%. This together with the desirable low optical
gaps, the favorable polymer miscibility with PC61BM and high
open-circuit voltages as well as high PCEs, demonstrates the
potential of this class of polymers as suitable donor candidates
for use in bottom subcells of efficient tandem devices.
A new set of BHJ PSCs were prepared, based on the four
polymers, by simply replacing PC61BM with PC71BM as the
acceptor (using the optimized processing conditions found for
PC61BM) in order to improve the photovoltaic performance by
enhancing the blend film absorption in the spectral range
between 350 and 550 nm for increased JSC.36 The resulting
photovoltaic performances are reported in Table 2, Table S5,
and Figure 3B. The JSC’s of optimized devices based on PMIM,
PT-MIM, or PBDTO-MIM combined with PC71BM were
slightly increased, while the VOC’s were the same or slightly
reduced. As a consequence, by replacing PC61BM with
PC71BM, the PCE remained almost identical for PMIM while
Unlike PBDTO-MIM, the best photovoltaic response of
PBDTA-MIM:PC61BM-based cell was obtained for as-cast
films. Indeed, a thermal annealing (110 °C 10 min) of the film
leads to a drop in JSC and PCE (similar to the case of PT-MIM,
Table S5), suggesting a slightly different self-aggregation
tendency despite their structural similarity. It is worth noting
that the PCEs for PBDTO-MIM and PBDTA-MIM together
with our previously reported PBDT-TIT and PBDT-OIO8l
follow the order PBDT-OIO (1.26%) < PBDT-TIT (4.22%) <
PBDTO-MIM (4.36%) < PBDTA-MIM (5.36%). This
H
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 4. UV−vis absorption spectra and EQE plots of optimized BHJ solar cells.
Figure 5. Tapping mode AFM images (5 μm × 5 μm) of the active layers from the optimized BHJ PSCs based on (A) 1:1 (w/w) PMIM:PC61BM
(RMS = ∼9.6 nm); (B) 1:1 (w/w) PT-MIM:PC61BM (RMS = ∼1.4 nm); (C) 1:1.5 (w/w) PBDTO-MIM:PC61BM (RMS = ∼3.2 nm); (D) 1:1.5
(w/w) PBDTA-MIM:PC61BM (RMS = ∼6.7 nm); (E) 1:1 (w/w) PMIM:PC71BM (RMS = ∼9.2 nm); (F) 1:1 (w/w) PT-MIM:PC71BM (RMS =
∼5.4 nm); (G) 1:1.5 (w/w) PBDTO-MIM:PC71BM (RMS = ∼2.3 nm); and (H) 1:1.5 (w/w) PBDTA-MIM:PC71BM (RMS = ∼6.1 nm).
for PT-MIM was increased from 3.14% to 3.55% and for
PBDTO-MIM from 4.36% to 4.98%. PBDTA-MIM:PC71BM-
based cells exhibit, compared to the analogous PC61BM-based
devices, a slight drop in JSC (from 10.20 to 9.47 mA/cm2) and
FF (from 65% to 61%). This may suggest, in these conditions, a
slightly poorer miscibility of PC71BM with this polymer,4q,36e
which partially hinders an optimal self-organization of the BHJ
components, as also confirmed by the reduced hole mobility of
PBDTA-MIM blended with PC71BM (1.2 × 10−4 cm2 V−1 s−1)
compared to the analogous PC61BM-based films (2.7 × 10−4
cm2 V−1 s−1). In order to gain insight into these differences,
deeper investigations and additional optimization processes are
underway since higher PCEs can be reached, in particular using
PBDTO-MIM and PBDTA-MIM as the donor and PC71BM as
the acceptor.
these devices were measured (Figure 4). The external quantum
efficiency (EQE) spectra of the optimized BHJ PSCs with
PC61BM or PC71BM as the acceptor (Figures 4B and 4D,
respectively) are consistent with the broad optical absorption
spectra of the active blends (Figures 4A and 4C, respectively).
The EQE spectra for the optimized cells using PC61BM as the
acceptor (Figure 4B) exhibit the highest ratio of photo-
generated electrons collected by the PSCs to incident photons
in the spectral range between 590 and 710 nm, corresponding
to the lower-energy absorption band of the polymers. Figure
4D shows the EQE responses of the most efficient PSCs using
PC71BM as the acceptor, in which the spectral profiles,
compared to the corresponding PC61BM-based devices, are
mainly enhanced in the wavelength range between 380 and 530
nm, in agreement with the absorption profile of PC71BM. As a
consequence, the EQE in this spectral region showed maximum
values of 30% (at 455 nm), 64% (at 430 nm), 59% (at 470 nm),
In order to enhance our understanding of the different
photocurrent generation efficiencies, the spectral responses of
I
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
and 46% (at 460 nm) respectively for PMIM-, PT-MIM-,
PBDTO-MIM-, and PBDTA-MIM-based PSCs with PC71BM
as the acceptor. Note that the red portion (>550 nm) of these
EQE profiles is slightly reduced as a possible consequence of
slightly different (induced) polymer self-organization or phase
separation. Convolution of these EQE spectra with the AM1.5
solar spectrum gave calculated JSC’s in good agreement, within a
∼10% experimental error, with those obtained from J−V
measurements (Table S6).
Morphological Studies. It is known that the nano-
morphology of the BHJ active layer plays an important role
in determining the photovoltaic performance.35a,b,37 The
morphological differences of the optimized polymer-based
films were analyzed by tapping-mode atomic force microscopy
(AFM). Figure 5 shows the topographic images of the films
from the optimized PSCs (reported in Table 2) based on these
polymers blended with either PC61BM or PC71BM.
The topographic image of PMIM:PC61BM (Figure 5A)
exhibits quite large and poorly defined domains, which may
suggest a suboptimal phase segregation of the BHJ
components. Thus, poorly formed percolation pathways lead
to limited charge transport and/or collection processes within
the active blends, as confirmed by the low EQE, JSC, FF, PCE,
and mobility obtained from the corresponding BHJ films. A
similar nanomorphology can be observed also for the PT-
MIM:PC61BM-based film (Figure 5B), which correlates well
with the low FF and charge mobility of the corresponding
device. This could be ascribed to the relatively low molar
masses and the relatively low D:A phase separation tendency of
PMIM and PT-MIM. On the other hand, the slightly reduced
domain sizes in PT-MIM:PC61BM-based film compared to
PMIM:PC61BM-based one correlates well with the higher JSC
and hole mobility of the former one.
On the other hand, the surface morphologies of PBDTO-
MIM:PC61BM- and PBDTA-MIM:PC61BM-based films (Fig-
ures 5C and 5D, respectively) are significantly different and
characterized by reduced domain sizes with finer nanostruc-
tures, likely due to the presence of BDT units that favors a
better intermolecular π−π stacking and ordering of the polymer
chains, which promotes a fine D:A intermixing with proper BHJ
nanoscale phase separation, reflected in the higher photovoltaic
performance of the resulting PBDTO-MIM- and PBDTA-
MIM-based solar cells.
CONCLUSIONS
■
In summary, we have designed and synthesized four donor−
acceptor conjugated alternating copolymers based on the same
isoindigo acceptor unit, which combine different electron-rich
thiophene-based donor moieties, in order to explore the
potential of isoindigo-based polymers for efficient donor
materials with high photovoltages in solar cells. We have
systematically investigated the influence of the different
electron-rich units on the structural, optical, electrochemical,
and photovoltaic properties. Mobility measurements, morpho-
logical studies, and quantum-chemical calculations including
DFT and TD-DFT were carried out to find out the insights
into the differences. We find that the PCEs of the resulting
PSCs can be improved by over 3-fold through a rational
structural modification of the donor moiety, which highlights
the importance of carefully choosing appropriate chemical
structure to design efficient donor−acceptor polymers for solar
cells. Among these polymers, PBDA-MIM exhibits a PCE of
5.4%, which is the highest value obtained for isoindigo-based
polymers for conventional BHJ PSCs combined with PC61BM
as the acceptor to date. This further emphasizes the use of
isoindigo as an effective acceptor unit for designing active
donor materials and demonstrates the potential of this class of
polymers as donor candidates for bottom cells in tandem
devices, which combine low optical gaps, promising efficiencies,
and desirable open-circuit voltages into isoindigo-based
polymers with PC61BM as the acceptor.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental details, device fabrication and characterization,
synthesis of monomers and polymers, square-wave voltammo-
grams, additional data of DFT and TD-DFT calculations,
details for optimizing processing conditions of solar cell
fabrication and calculated short current densities for compar-
ison of optimized cells, and abbreviations. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Authors
*E-mail wenliu@chalmers.se (W.Z.).
*E-mail mats.andersson@chalmers.se (M.R.A.).
Notes
■
The AFM images of the analogous blends with PC71BM
(Figures 5E−H) reveal a nearly identical trend already observed
for PC61BM-based films without evident and/or significant
variations, in agreement with the comparable electrical
performance. Note that the PT-MIM:PC71BM film (Figure
5F) exhibits more defined and structured domains which
correlate well with the relatively increased FF and PCE. As
mentioned above, the well-organized morphologies and the
relatively high charge mobilities measured on PBDTO-MIM
and PBDTA-MIM films confirm also in this case the formation
of optimal nanoscale phase separation and interpenetrating
networks in the entire photoactive layer, in agreement with the
high FF and PCEs up to ∼5% of these devices.
Finally, note that PBDTA-MIM-based films, independently
from the acceptor material employed, exhibit well-organized
and fine nanostructured motifs, highlighting an optimal self-
organization and phase segregation tendency in BHJ layers,
thus resulting in optimal electrical properties, with PCEs of
∼5%, without the need of additional thermal treatments.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was financially supported by the European
Commission FP7 collaborative project SUNFLOWER (FP7-
ICT-2011-7, Grant No. 287594). W.Z. also thanks the Swedish
Energy Agency for financial support. We also acknowledge the
Swedish Research Council and Chalmers’ Area of Advance
Materials Science for funding. The quantum chemical
computations were performed on resources provided by the
Swedish National Infrastructure for Computing (SNIC) at
C3SE in Gothenburg. Anders Mårtensson was gratefully
acknowledged for GPC characterizations and Timothy Steckler
for fruitful discussions.
REFERENCES
■
(1) (a) Li, G.; Zhu, R.; Yang, Y. Nat. Photonics 2012, 6, 153−161.
(b) Brabec, C. J. Sol. Energy Mater. Sol. Cells 2004, 83, 273−292.
(c) Wang, E.; Wang, L.; Lan, L.; Luo, C.; Zhuang, W.; Peng, J.; Cao, Y.
J
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Appl. Phys. Lett. 2008, 92, 033307. (d) Zhen, H. Y.; Li, K.; Huang, Z.
Y.; Tang, Z.; Wu, R. M.; Li, G. L.; Liu, X.; Zhang, F. L. Appl. Phys. Lett.
2012, 100, 213901. (e) Bolognesi, M.; Tessarolo, M.; Posati, T.;
Nocchetti, M.; Benfenati, V.; Seri, M.; Ruani, G.; Muccini, M. Org.
Photon. Photovolt. 2013, 1, 1−10.
Ma, Z. F.; Zhang, Z.; Vandewal, K.; Henriksson, P.; Inganas, O.;
̈
Zhang, F. L.; Andersson, M. R. J. Am. Chem. Soc. 2011, 133, 14244−
14247. (j) Ma, Z. F.; Wang, E. G.; Vandewal, K.; Andersson, M. R.;
Zhang, F. L. Appl. Phys. Lett. 2011, 99, 143302. (k) Mei, J.; Kim, D. H.;
Ayzner, A. L.; Toney, M. F.; Bao, Z. J. Am. Chem. Soc. 2011, 133,
20130−20133. (l) Ma, Z. F.; Wang, E. G.; Jarvid, M. E.; Henriksson,
(2) Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. Science
1995, 270, 1789−1791.
P.; Inganas, O.; Zhang, F. L.; Andersson, M. R. J. Mater. Chem. 2012,
̈
(3) (a) Service, R. F. Science 2011, 332, 293. (b) He, Z. C.; Zhong, C.
M.; Su, S. J.; Xu, M.; Wu, H. B.; Cao, Y. Nat. Photonics 2012, 6, 591−
595. (c) Small, C. E.; Chen, S.; Subbiah, J.; Amb, C. M.; Tsang, S.-W.;
Lai, T.-H.; Reynolds, J. R.; So, F. Nat. Photonics 2011, 6, 115−120.
(d) McCulloch, I. Adv. Mater. 2013, 25, 1811−1812.
22, 2306−2314.
(9) Wahl, A.; Bagard, P. Bull. Soc. Chim. Fr. 1910, 5, 1039−1043.
(10) (a) Hu, C.; Fu, Y.; Li, S.; Xie, Z.; Zhang, Q. Polym. Chem. 2012,
3, 2949−2955. (b) Stalder, R.; Grand, C.; Subbiah, J.; So, F.; Reynolds,
J. R. Polym. Chem. 2012, 3, 89−92. (c) Vandewal, K.; Ma, Z. F.;
Bergqvist, J.; Tang, Z.; Wang, E. G.; Henriksson, P.; Tvingstedt, K.;
(4) (a) Chen, H. Y.; Hou, J. H.; Zhang, S. Q.; Liang, Y. Y.; Yang, G.
W.; Yang, Y.; Yu, L. P.; Wu, Y.; Li, G. Nat. Photonics 2009, 3, 649−653.
(b) Liang, Y. Y.; Feng, D. Q.; Wu, Y.; Tsai, S. T.; Li, G.; Ray, C.; Yu, L.
P. J. Am. Chem. Soc. 2009, 131, 7792−7799. (c) Liang, Y. Y.; Xu, Z.;
Xia, J. B.; Tsai, S. T.; Wu, Y.; Li, G.; Ray, C.; Yu, L. P. Adv. Mater.
2010, 22, E135−E138. (d) Huo, L.; Zhang, S.; Guo, X.; Xu, F.; Li, Y.;
Hou, J. Angew. Chem., Int. Ed. 2011, 50, 9697−9702. (e) Zhou, H. X.;
Yang, L. Q.; Stuart, A. C.; Price, S. C.; Liu, S. B.; You, W. Angew.
Chem., Int. Ed. 2011, 50, 2995−2998. (f) Price, S. C.; Stuart, A. C.;
Yang, L.; Zhou, H.; You, W. J. Am. Chem. Soc. 2011, 133, 4625−4631.
(g) Huang, Y.; Guo, X.; Liu, F.; Huo, L.; Chen, Y.; Russell, T. P.; Han,
C. C.; Li, Y.; Hou, J. Adv. Mater. 2012, 24, 3383−3389. (h) Piliego, C.;
Holcombe, T. W.; Douglas, J. D.; Woo, C. H.; Beaujuge, P. M.;
Frechet, J. M. J. J. Am. Chem. Soc. 2010, 132, 7595−7597. (i) Chu, T.-
Andersson, M. R.; Zhang, F. L.; Inganas, O. Adv. Funct. Mater. 2012,
̈
22, 3480−3490. (d) Cao, K.; Wu, Z.; Li, S.; Sun, B.; Zhang, G.; Zhang,
Q. J. Polym. Sci., Part A: Polym. Chem. 2013, 51, 94−100. (e) Xu, X. F.;
Cai, P.; Lu, Y.; Choon, N. S.; Chen, J. W.; Hu, X.; Ong, B. S. J. Polym.
Sci., Part A: Polym. Chem. 2012, 51, 424−434.
(11) Ko, S.; Hoke, E. T.; Pandey, L.; Hong, S.; Mondal, R.; Risko, C.;
Yi, Y.; Noriega, R.; McGehee, M. D.; Bred
Am. Chem. Soc. 2012, 134, 5222−5232.
́
as, J.-L.; Salleo, A.; Bao, Z. J.
(12) Aleman, C.; Julia, L. J. Phys. Chem. 1996, 100, 1524−1529.
(13) (a) Barbarella, G.; Zambianchi, M.; Bongini, A.; Antolini, L. Adv.
Mater. 1992, 4, 282−285. (b) Fichou, D. J. Mater. Chem. 2000, 10,
571−588.
(14) (a) Hou, J. H.; Chen, H. Y.; Zhang, S. Q.; Chen, R. I.; Yang, Y.;
Wu, Y.; Li, G. J. Am. Chem. Soc. 2009, 131, 15586−15587. (b) Li, Z.;
Lu, J. P.; Tse, S. C.; Zhou, J. Y.; Du, X. M.; Tao, Y.; Ding, J. F. J. Mater.
Chem. 2011, 21, 3226−3233. (c) Shi, C. J.; Yao, Y.; Yang, Y.; Pei, Q. B.
J. Am. Chem. Soc. 2006, 128, 8980−8986.
́
Y.; Lu, J.; Beaupre, S.; Zhang, Y.; Pouliot, J.-R.; Wakim, S.; Zhou, J.;
Leclerc, M.; Li, Z.; Ding, J.; Tao, Y. J. Am. Chem. Soc. 2011, 133,
4250−4253. (j) Amb, C. M.; Chen, S.; Graham, K. R.; Subbiah, J.;
Small, C. E.; So, F.; Reynolds, J. R. J. Am. Chem. Soc. 2011, 133,
10062−10065. (k) Chen, H.-C.; Chen, Y.-H.; Liu, C.-C.; Chien, Y.-C.;
Chou, S.-W.; Chou, P.-T. Chem. Mater. 2012, 24, 4766−4772.
(l) Chang, C. Y.; Cheng, Y. J.; Hung, S. H.; Wu, J. S.; Kao, W. S.; Lee,
C. H.; Hsu, C. S. Adv. Mater. 2012, 24, 549−553. (m) Xu, Y.-X.;
Chueh, C.-C.; Yip, H.-L.; Ding, F.-Z.; Li, Y.-X.; Li, C.-Z.; Li, X.; Chen,
W.-C.; Jen, A. K. Y. Adv. Mater. 2012, 24, 6356−6361. (n) Guo, X.;
Zhang, M.; Tan, J.; Zhang, S.; Huo, L.; Hu, W.; Li, Y.; Hou, J. Adv.
Mater. 2012, 24, 6536−6541. (o) Son, H. J.; Lu, L.; Chen, W.; Xu, T.;
Zheng, T.; Carsten, B.; Strzalka, J.; Darling, S. B.; Chen, L. X.; Yu, L.
Adv. Mater. 2012, 25, 838−843. (p) Dou, L.; Chang, W.-H.; Gao, J.;
Chen, C.-C.; You, J.; Yang, Y. Adv. Mater. 2012, 25, 825−831. (q) Su,
M. S.; Kuo, C. Y.; Yuan, M. C.; Jeng, U. S.; Su, C. J.; Wei, K. H. Adv.
Mater. 2011, 23, 3315−3319.
(15) (a) Quillinan, A. J.; Khan, E. A.; Scheinmann, F. J. Chem. Soc.,
Chem. Commun. 1974, 1030−1031. (b) Kobayashi, Y.; Murugesh, M.
G.; Nakano, M.; Takahisa, E.; Usmani, S. B.; Ainai, T. J. Org. Chem.
2002, 67, 7110−7123. (c) Slocum, D. W.; Gierer, P. L. J. Org. Chem.
1976, 41, 3668−3673. (d) Beimling, P.; Koβmehl, G. Chem. Ber. 1986,
119, 3198−3203. (e) Anthony, J. E.; Brooks, J. S.; Eaton, D. L.; Parkin,
S. R. J. Am. Chem. Soc. 2001, 123, 9482−9483.
(16) (a) Xiang, N.; Huang, X.; Feng, X.; Liu, Y.; Zhao, B.; Deng, L.;
Shen, P.; Fei, J.; Tan, S. Dyes Pigm. 2011, 88, 75−83. (b) Cai, T. Q.;
Zhou, Y.; Wang, E. G.; Hellstrom, S.; Zhang, F. L.; Xu, S. A.; Inganas,
̈
̈
O.; Andersson, M. R. Sol. Energy Mater. Sol. Cells 2010, 94, 1275−
1281.
(17) Smith, K.; Barratt, M. L. J. Org. Chem. 2007, 72, 1031−1034.
(18) Bao, Z. N.; Chan, W. K.; Yu, L. P. J. Am. Chem. Soc. 1995, 117,
12426−12435.
(5) Havinga, E. E.; ten Hoeve, W.; Wynberg, H. Polym. Bull. 1992,
29, 119−126.
(6) (a) Roncali, J. Chem. Rev. 1997, 97, 173−206. (b) Scharber, M.
C.; Wuhlbacher, D.; Koppe, M.; Denk, P.; Waldauf, C.; Heeger, A. J.;
Brabec, C. L. Adv. Mater. 2006, 18, 789−794. (c) Bredas, J. L.; Norton,
J. E.; Cornil, J.; Coropceanu, V. Acc. Chem. Res. 2009, 42, 1691−1699.
(d) Zhou, H. X.; Yang, L. Q.; Stoneking, S.; You, W. ACS Appl. Mater.
Interfaces 2010, 2, 1377−1383. (e) Zhou, H.; Yang, L.; You, W.
Macromolecules 2012, 45, 607−632.
(19) Lei, T.; Cao, Y.; Zhou, X.; Peng, Y.; Bian, J.; Pei, J. Chem. Mater.
2012, 24, 1762−1770.
(20) Wegner, G. Thin Solid Films 1992, 216, 105−116.
(21) Muller, C.; Wang, E.; Andersson, L. M.; Tvingstedt, K.; Zhou,
̈
Y.; Andersson, M. R.; Inganas, O. Adv. Funct. Mater. 2010, 20, 2124−
̈
2131.
(22) Risko, C.; McGehee, M. D.; Bredas, J. L. Chem. Sci. 2011, 2,
1200−1218.
(7) (a) Dennler, G.; Scharber, M. C.; Brabec, C. J. Adv. Mater. 2009,
21, 1323−1338. (b) Dennler, G.; Scharber, M. C.; Ameri, T.; Denk, P.;
Forberich, K.; Waldauf, C.; Brabec, C. J. Adv. Mater. 2008, 20, 579−
583.
(23) (a) Yuan, M.-C.; Chiu, M.-Y.; Chiang, C.-M.; Wei, K.-H.
Macromolecules 2010, 43, 6270−6277. (b) Guo, X.; Zhang, M. J.; Huo,
L. J.; Xu, F.; Wu, Y.; Hou, J. H. J. Mater. Chem. 2012, 22, 21024−
21031. (c) Peet, J.; Kim, J. Y.; Coates, N. E.; Ma, W. L.; Moses, D.;
Heeger, A. J.; Bazan, G. C. Nat. Mater. 2007, 6, 497−500.
(d) Muhlbacher, D.; Scharber, M.; Morana, M.; Zhu, Z. G.; Waller,
D.; Gaudiana, R.; Brabec, C. Adv. Mater. 2006, 18, 2884.
(24) Lei, T.; Dou, J.-H.; Pei, J. Adv. Mater. 2012, 24, 6457−6461.
(8) (a) Mei, J. G.; Graham, K. R.; Stalder, R.; Reynolds, J. R. Org.
Lett. 2010, 12, 660−663. (b) Stalder, R.; Mei, J.; Reynolds, J. R.
Macromolecules 2010, 43, 8348−8352. (c) Zhang, G.; Fu, Y.; Xie, Z.;
Zhang, Q. Macromolecules 2011, 44, 1414−1420. (d) Wang, E. G.; Ma,
Z. F.; Zhang, Z.; Henriksson, P.; Inganas, O.; Zhang, F. L.; Andersson,
̈
M. R. Chem. Commun. 2011, 47, 4908−4910. (e) Liu, B.; Zou, Y. P.;
Peng, B.; Zhao, B.; Huang, K. L.; He, Y. H.; Pan, C. Y. Polym. Chem
2011, 2, 1156−1162. (f) Graham, K.; Mei, J.; Stalder, R.; Shim, J.;
Cheun, H.; So, F.; Steffy, F.; Kippelen, B.; Reynolds, J. R. ACS Appl.
Mater. Interfaces 2011, 3, 1210−1215. (g) Lei, T.; Cao, Y.; Fan, Y.; Liu,
C.-J.; Yuan, S.-C.; Pei, J. J. Am. Chem. Soc. 2011, 133, 6099−6101.
(h) Stalder, R.; Mei, J.; Subbiah, J.; Grand, C.; Estrada, L. A.; So, F.;
Reynolds, J. R. Macromolecules 2011, 44, 6303−6310. (i) Wang, E. G.;
(25) (a) Wang, E.; Hou, L.; Wang, Z.; Hellstrom, S.; Mammo, W.;
̈
Zhang, F.; Inganas, O.; Andersson, M. R. Org. Lett. 2010, 12, 4470−
̈
4473. (b) Wang, E.; Hou, L.; Wang, Z.; Ma, Z.; Hellstrom, S.; Zhuang,
̈
W.; Zhang, F.; Inganas, O.; Andersson, M. R. Macromolecules 2011, 44,
̈
2067−2073.
(26) (a) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. Rev. B 1988, 37,
785−789. (b) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.
(c) Hehre, W. J.; Ditchfie, R.; Pople, J. A. J. Chem. Phys. 1972, 56,
K
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
2257−2261. (d) Harihara, P. C.; Pople, J. A. Theor. Chim. Acta 1973,
28, 213−222.
(27) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Keith, T.; Kobayashi, R.; Normand, J.;
Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.;
Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.;
Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.;
Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador,
P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
Revision B.01; Gaussian, Inc.: Wallingford, CT, 2010.
(28) Salvatori, P.; Mosconi, E.; Wang, E.; Andersson, M. R.; Muccini,
M.; De Angelis, F. J. Phys. Chem. C 2013, 117, 17940−17954.
(29) Chang, H.-H.; Tsai, C.-E.; Lai, Y.-Y.; Chiou, D.-Y.; Hsu, S.-L.;
Hsu, C.-S.; Cheng, Y.-J. Macromolecules 2012, 45, 9282−9291.
(30) Casida, M. E. J. Mol. Struct.: THEOCHEM 2009, 914, 3−18.
(31) O’Boyle, N. M.; Tenderholt, A. L.; Langner, K. M. J. Comput.
Chem. 2008, 29, 839−845.
(32) Brabec, C. J.; Dyakonov, V.; Parisi, J.; Sariciftci, N. S. Organic
Photovoltaics: Concepts and Realization; Springer-Verlag: Berlin, 2003;
Vol. 60.
(33) (a) Lee, J. K.; Ma, W. L.; Brabec, C. J.; Yuen, J.; Moon, J. S.;
Kim, J. Y.; Lee, K.; Bazan, G. C.; Heeger, A. J. J. Am. Chem. Soc. 2008,
130, 3619−3623. (b) Liu, X. L.; Huettner, S.; Rong, Z. X.; Sommer,
M.; Friend, R. H. Adv. Mater. 2012, 24, 669−674. (c) Rogers, J. T.;
Schmidt, K.; Toney, M. F.; Kramer, E. J.; Bazan, G. C. Adv. Mater.
2011, 23, 2284−2288.
(34) Brabec, C. J.; Cravino, A.; Meissner, D.; Sariciftci, N. S.;
Fromherz, T.; Rispens, M. T.; Sanchez, L.; Hummelen, J. C. Adv.
Funct. Mater. 2001, 11, 374−380.
(35) (a) Perez, M. D.; Borek, C.; Forrest, S. R.; Thompson, M. E. J.
Am. Chem. Soc. 2009, 131, 9281−9286. (b) Stevens, D. M.; Speros, J.
C.; Hillmyer, M. A.; Frisbie, C. D. J. Phys. Chem. C 2011, 115, 20806−
20816. (c) Liu, J.; Shi, Y. J.; Yang, Y. Adv. Funct. Mater. 2001, 11, 420−
424.
(36) (a) Wienk, M. M.; Kroon, J. M.; Verhees, W. J. H.; Knol, J.;
Hummelen, J. C.; van Hal, P. A.; Janssen, R. A. J. Angew. Chem., Int. Ed.
2003, 42, 3371−3375. (b) Morana, M.; Wegscheider, M.; Bonanni, A.;
Kopidakis, N.; Shaheen, S.; Scharber, M.; Zhu, Z.; Waller, D.;
Gaudiana, R.; Brabec, C. Adv. Funct. Mater. 2008, 18, 1757−1766.
(c) Sariciftci, N. S.; Smilowitz, L.; Heeger, A. J.; Wudl, F. Science 1992,
258, 1474−1476. (d) Hummelen, J. C.; Knight, B. W.; LePeq, F.;
Wudl, F.; Yao, J.; Wilkins, C. L. J. Org. Chem. 1995, 60, 532−538.
(e) Zhang, F.; Zhuo, Z.; Zhang, J.; Wang, X.; Xu, X.; Wang, Z.; Xin, Y.;
Wang, J.; Wang, J.; Tang, W.; Xu, Z.; Wang, Y. Sol. Energy Mater. Sol.
Cells 2012, 97, 71−77.
(37) (a) Dang, M. T.; Hirsch, L.; Wantz, G.; Wuest, J. D. Chem. Rev.
2013, 113, 3734−3765. (b) Moule, A. J.; Meerholz, K. Adv. Funct.
Mater. 2009, 19, 3028−3036.
L
dx.doi.org/10.1021/ma401691r | Macromolecules XXXX, XXX, XXX−XXX